329 Can ChatGPT Deliver the Perfect Sales Script?

THE Sales Japan Series by Dale Carnegie Training Tokyo Japan

In many sales organisations, sales scripts are a way of harmonising the messaging as well as getting newbies up to speed. ChatGPT can come up with a sales script within about one second. We just need to design the question we want answered intelligently and it will do all of the work. When you have to listen to sales people using sales scripts on you, the conclusion is ChatGPT must surely be able to do a better job, because the bar is set so low. Of course there are scripts and there are scripts, so the breadth of possibility is vast.

Even the best designed script will fail in the hands of idiot sales managers. I blame the managers for the sins of their staff, who have to deliver these scripts and do so in a pathetic fashion. When you can hear the cadence is someone reading the script to you, then you know you are dealing with an idiot-led organisation. If the manager was doing their job then this would never happen, but sadly it does and it negates any potential the script had to identify business.

The other challenge with scripts, be they AI generated or not, is to match the content with the client’s situation. Hopefully, we are going to try and join the conversation going on in the mind of the buyer with our first approach. “How are you today” is a pet peeve of mine as an opening volley in cold calling. As soon as I hear that line my toes curl up and I brace myself. I know what comes next won’t be intelligently constructed or useful to me. I have problems. What can this caller do to help me with my problems?

When we programme ChatGPT properly to work on the various problems the buyer may be having, then we can get a range of scripts to suit the situation. The list of problems is not hard to put together – lack of revenues, costs too high, can’t get or keep staff, stiff competition in the market driving down pricing, not enough differentiation in our solution, the buyer is happy with the current supplier, etc.

We will have to work a bit more on the scripts to personalise them for each buyer, but we can get a start using the machine. The other key part is the delivery, because the buyer doesn’t get their side of the script to follow and they can be unpredictable regarding what they might say. As salespeople, we have to be able to pivot on the spot and deal with whatever the buyer sends towards us. ChatGPT is of no help when we are live and the process is fluid.

An area it may be able to help is the designing of the solution explanation. When we get a chance to talk about what we can do for the buyer, we usually start with the details of the solution – the nitty gritty. Some of this nitty gritty is of more value than other parts and we should be seeking help to design for that rich vein of information which will delight the buyer. The next stage up is to outline the benefits to the buyer and this is where the machine helps us to get a nice list of these and we can use this as the base from which to select elements to impress the buyer. We have to feed it the information first though, so if we know this stuff, then why do we need the machine to re-work it for us? Can’t we just use what we have come up with?

The organisation of the evidence could use some ChatGPT love. Ideally, if we can do it we can organise our data based on the industry sub-set and the solution chosen and then match these with a satisfied client who took our solution, executed on it and reaped the rewards of doing so. Once we think about this structure the complexity increases very quickly. That is where programming the machine to absorb all of this data and then spit it back out in a tangible format can be helpful. We have to command all of this data at our fingertips because in the to and fro with the buyer, all of this will have to be coming out of our heads. The machine will consolidate the information, but we still have to absorb it and be ready to draw on the relevant parts, for the buyer sitting in front of us.

There is always a high degree of personalisation of the messaging for the buyers and no machine will be able to handle that requirement in real time. The base of the information pool or the examples or the alternatives are certainly things we can always use help with. We should always keep in mind that just because we have the ability to tap this fountain of wisdom and knowledge, that we shouldn’t always do so. The buyer should be talking 80% of the time and we just 20%. The temptation is for ChatGPT to turn into a professor and make us “talkers” rather than “questioners”. Introducing who we are and what we do and explaining how the solution works are three excellent occasions for us to be talking and the machine can help us out here.

Like all technology the intelligence is in what we feed into the machine. This is where we have to try and educate it on addressing what we want for our buyers. This takes practice but the upside is the speed of production is basically instant, so we can invest time to keep polishing the input questions.